RE: RAID system max throughput
Thanks to everyone that shared their thoughts on this issue. Steven - Thanks very much for your insight of throughput vs. latency. I think this reduces the concepts to a simple enough level that I can explain them to others. The system administrators also do our wide-area networks and the focus tends to be on throughput, but they have also realized that latency times can also produce poor performance. Dennis Williams DBA Lifetouch, Inc. [EMAIL PROTECTED] -Original Message- Sent: Friday, December 07, 2001 8:40 PM To: Multiple recipients of list ORACLE-L -- DENNIS WILLIAMS [EMAIL PROTECTED] Whenever I discuss disk waits with my system administrator, I always get the reply that the RAID system isn't anywhere near its rated throughput. Maybe I'm wrong, but I don't see any of the tuning books mentioning that as a relevant performance characteristic. However, I've never been able to move the discussion beyond this point. Can anyone straighten me out on this point or point me to a resource that might be applicable. Our system is Oracle 8.1.6, Compaq Tru64. We use hardware RAID-5 with a battery-backed RAM cache, and have about 3 RAID sets (plus some extra disks for redo logs, etc.), and performance is fine, but I'm always looking as to how we can improve Oracle performance. The application is our corporate ERP system. Two things will zap you on device I/O: bandwidth or latency. Most people look at bandwidth -- same basic numbers for both networking and disks. Latency is basically the turnaround time. If you screw up the setup of any I/O system then latency can reduce performance to the point where bandwidth is irrelevant. By analogy, you can put concrete tires on a Porsche and go nowhere also. The real measure of what's going on starts at the O/S level looking at the frequency and duration of proc's in a device wait state (a.k.a. blocked for I/O) on the disks. If this is minimal then forget it. You can also end up with screwy results on large shared disk systems due to competition. SAN's can get placed on overloaded network segments; ERP's can easily get hot-spots from various users colliding. In general RAID5 with a stripe size == system I/O page will perform rather nicely. If your system page sizes vary or the raidset has an offball number of disks (e.g., 6 drives for an 8K page) then you'll take a hit writing extra data to maintain the RAID5 parity. -- Steven Lembark 2930 W. Palmer Workhorse Computing Chicago, IL 60647 +1 800 762 1582 -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: Steven Lembark INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: DENNIS WILLIAMS INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: RAID system max throughput
I'm not sure that I understand the question. Is it: A. Under max load the RAID does not perform up to its specs? or B. Oracle does not stress the RAID enough for it to reach its max performance. If A, then have the RAID vendor fix whatever the problem is. If B, then good for you. You have a well designed system. DENNIS WILLIAMS To: Multiple recipients of list ORACLE-L DWILLIAMS [EMAIL PROTECTED] @LIFETOUCH.COcc: M Subject: RAID system max throughput Sent by: root 12/07/2001 11:40 AM Please respond to ORACLE-L Whenever I discuss disk waits with my system administrator, I always get the reply that the RAID system isn't anywhere near its rated throughput. Maybe I'm wrong, but I don't see any of the tuning books mentioning that as a relevant performance characteristic. However, I've never been able to move the discussion beyond this point. Can anyone straighten me out on this point or point me to a resource that might be applicable. Our system is Oracle 8.1.6, Compaq Tru64. We use hardware RAID-5 with a battery-backed RAM cache, and have about 3 RAID sets (plus some extra disks for redo logs, etc.), and performance is fine, but I'm always looking as to how we can improve Oracle performance. The application is our corporate ERP system. Dennis Williams DBA Lifetouch, Inc. [EMAIL PROTECTED] -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: DENNIS WILLIAMS INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
RE: RAID system max throughput
tday6 - Thanks for replying. No, the RAID seems to perform fine. I can perform certain Oracle operations (usually DBA activities) and kick the RAID I/O really high (considerably below the vendor specs). In normal usage the I/O is far below this level. Statspack analysis shows disk waits to be our greatest wait. Does this mean that I have a superbly tuned system and there is nothing more to do? If I can get an amen to that then I can leave happy for the weekend. Part of what concerned me was that none of the Oracle tuning books I have mentions vendor specs for RAID I/O at all. Thanks for your ideas. Dennis Williams DBA Lifetouch, Inc. [EMAIL PROTECTED] -Original Message- Sent: Friday, December 07, 2001 12:13 PM To: Multiple recipients of list ORACLE-L I'm not sure that I understand the question. Is it: A. Under max load the RAID does not perform up to its specs? or B. Oracle does not stress the RAID enough for it to reach its max performance. If A, then have the RAID vendor fix whatever the problem is. If B, then good for you. You have a well designed system. DENNIS WILLIAMS To: Multiple recipients of list ORACLE-L DWILLIAMS [EMAIL PROTECTED] @LIFETOUCH.COcc: M Subject: RAID system max throughput Sent by: root 12/07/2001 11:40 AM Please respond to ORACLE-L Whenever I discuss disk waits with my system administrator, I always get the reply that the RAID system isn't anywhere near its rated throughput. Maybe I'm wrong, but I don't see any of the tuning books mentioning that as a relevant performance characteristic. However, I've never been able to move the discussion beyond this point. Can anyone straighten me out on this point or point me to a resource that might be applicable. Our system is Oracle 8.1.6, Compaq Tru64. We use hardware RAID-5 with a battery-backed RAM cache, and have about 3 RAID sets (plus some extra disks for redo logs, etc.), and performance is fine, but I'm always looking as to how we can improve Oracle performance. The application is our corporate ERP system. Dennis Williams DBA Lifetouch, Inc. [EMAIL PROTECTED] -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: DENNIS WILLIAMS INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: DENNIS WILLIAMS INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
RE: RAID system max throughput
Dennis, I'm no RAID guru, but I can sure imagine disk heads thrashing around, trying to satisfy a mix of sequential and random reads and writes, causing the DB to wait, but not getting anywhere near the rated throughput for the RAID controller or channel. Could that possibly be the case? Jack Jack C. Applewhite Database Administrator/Developer OCP Oracle8 DBA iNetProfit, Inc. Austin, Texas www.iNetProfit.com [EMAIL PROTECTED] (512)327-9068 -Original Message- WILLIAMS Sent: Friday, December 07, 2001 10:40 AM To: Multiple recipients of list ORACLE-L Whenever I discuss disk waits with my system administrator, I always get the reply that the RAID system isn't anywhere near its rated throughput. Maybe I'm wrong, but I don't see any of the tuning books mentioning that as a relevant performance characteristic. However, I've never been able to move the discussion beyond this point. Can anyone straighten me out on this point or point me to a resource that might be applicable. Our system is Oracle 8.1.6, Compaq Tru64. We use hardware RAID-5 with a battery-backed RAM cache, and have about 3 RAID sets (plus some extra disks for redo logs, etc.), and performance is fine, but I'm always looking as to how we can improve Oracle performance. The application is our corporate ERP system. Dennis Williams DBA Lifetouch, Inc. [EMAIL PROTECTED] -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: Jack C. Applewhite INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
RE: RAID system max throughput
Can you get to the max throughput outside Oracle - for example, a straight read or write test. If you can, then it would reasonable to assume that your db could possibly get some more io throughput squeezed out of it. If the read or write test cannot get up to the claimed max, then maybe its time to ask for an explanation from the raid vendor. hth connor --- DENNIS WILLIAMS [EMAIL PROTECTED] wrote: tday6 - Thanks for replying. No, the RAID seems to perform fine. I can perform certain Oracle operations (usually DBA activities) and kick the RAID I/O really high (considerably below the vendor specs). In normal usage the I/O is far below this level. Statspack analysis shows disk waits to be our greatest wait. Does this mean that I have a superbly tuned system and there is nothing more to do? If I can get an amen to that then I can leave happy for the weekend. Part of what concerned me was that none of the Oracle tuning books I have mentions vendor specs for RAID I/O at all. Thanks for your ideas. Dennis Williams DBA Lifetouch, Inc. [EMAIL PROTECTED] -Original Message- Sent: Friday, December 07, 2001 12:13 PM To: Multiple recipients of list ORACLE-L I'm not sure that I understand the question. Is it: A. Under max load the RAID does not perform up to its specs? or B. Oracle does not stress the RAID enough for it to reach its max performance. If A, then have the RAID vendor fix whatever the problem is. If B, then good for you. You have a well designed system. DENNIS WILLIAMS To: Multiple recipients of list ORACLE-L DWILLIAMS [EMAIL PROTECTED] @LIFETOUCH.COcc: M Subject: RAID system max throughput Sent by: root 12/07/2001 11:40 AM Please respond to ORACLE-L Whenever I discuss disk waits with my system administrator, I always get the reply that the RAID system isn't anywhere near its rated throughput. Maybe I'm wrong, but I don't see any of the tuning books mentioning that as a relevant performance characteristic. However, I've never been able to move the discussion beyond this point. Can anyone straighten me out on this point or point me to a resource that might be applicable. Our system is Oracle 8.1.6, Compaq Tru64. We use hardware RAID-5 with a battery-backed RAM cache, and have about 3 RAID sets (plus some extra disks for redo logs, etc.), and performance is fine, but I'm always looking as to how we can improve Oracle performance. The application is our corporate ERP system. Dennis Williams DBA Lifetouch, Inc. [EMAIL PROTECTED] -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: DENNIS WILLIAMS INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: DENNIS WILLIAMS INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). = Connor McDonald
RE: RAID system max throughput
Jack - Well, that's what I thought. I could see where the disk would be a lot better about streaming data off the disk if the data was arranged in a favorable manner rather than randomly located. However, I was told that was simplistic thinking and that modern RAID systems are much more sophisticated than that. And I'm willing to concede that a RAID system is more complex than simple drives. I'm just hoping that someone on this list has more experience on the database/hardware interface. Thanks. -Original Message- Sent: Friday, December 07, 2001 3:25 PM To: Multiple recipients of list ORACLE-L Dennis, I'm no RAID guru, but I can sure imagine disk heads thrashing around, trying to satisfy a mix of sequential and random reads and writes, causing the DB to wait, but not getting anywhere near the rated throughput for the RAID controller or channel. Could that possibly be the case? Jack Jack C. Applewhite Database Administrator/Developer OCP Oracle8 DBA iNetProfit, Inc. Austin, Texas www.iNetProfit.com [EMAIL PROTECTED] (512)327-9068 -Original Message- WILLIAMS Sent: Friday, December 07, 2001 10:40 AM To: Multiple recipients of list ORACLE-L Whenever I discuss disk waits with my system administrator, I always get the reply that the RAID system isn't anywhere near its rated throughput. Maybe I'm wrong, but I don't see any of the tuning books mentioning that as a relevant performance characteristic. However, I've never been able to move the discussion beyond this point. Can anyone straighten me out on this point or point me to a resource that might be applicable. Our system is Oracle 8.1.6, Compaq Tru64. We use hardware RAID-5 with a battery-backed RAM cache, and have about 3 RAID sets (plus some extra disks for redo logs, etc.), and performance is fine, but I'm always looking as to how we can improve Oracle performance. The application is our corporate ERP system. Dennis Williams DBA Lifetouch, Inc. [EMAIL PROTECTED] -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: Jack C. Applewhite INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: DENNIS WILLIAMS INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: RAID system max throughput
Whenever someone (especially a vendor) says something like Don't worry about it..., I worry about it. Who told you that this was simplistic thinking? I've been told similar things a number of times - and proved them wrong in every single case. With hardware RAID, RAID-5 is just as fast as RAID 1+0. With EMC Symmetrix you don't want to stripe. SAME - Just splatter all your files randomly across a monster stripe set using every possible disk. And the ever popular one you are encountering now. A lot of those things are at true - to a point. Beyond that point it matters. Hardware RAID, cache, and such can buy you performance, but there is still some threshold beyond which the old-timey DBA intelligent file placement, striping and such will be necessary. There is a difference between good enough for now and optimal. I would rather build it better from the start, even if I don't need the performance immediately, than wait until its a crisis and only then frantically rebuild everything. See Gaja's paper on RAID at http://www.quest.com/whitepapers/Raid1.pdf . -Don Granaman [OraSaurus] - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Friday, December 07, 2001 4:31 PM Jack - Well, that's what I thought. I could see where the disk would be a lot better about streaming data off the disk if the data was arranged in a favorable manner rather than randomly located. However, I was told that was simplistic thinking and that modern RAID systems are much more sophisticated than that. And I'm willing to concede that a RAID system is more complex than simple drives. I'm just hoping that someone on this list has more experience on the database/hardware interface. Thanks. -Original Message- Sent: Friday, December 07, 2001 3:25 PM To: Multiple recipients of list ORACLE-L Dennis, I'm no RAID guru, but I can sure imagine disk heads thrashing around, trying to satisfy a mix of sequential and random reads and writes, causing the DB to wait, but not getting anywhere near the rated throughput for the RAID controller or channel. Could that possibly be the case? Jack Jack C. Applewhite Database Administrator/Developer OCP Oracle8 DBA iNetProfit, Inc. Austin, Texas www.iNetProfit.com [EMAIL PROTECTED] (512)327-9068 -Original Message- WILLIAMS Sent: Friday, December 07, 2001 10:40 AM To: Multiple recipients of list ORACLE-L Whenever I discuss disk waits with my system administrator, I always get the reply that the RAID system isn't anywhere near its rated throughput. Maybe I'm wrong, but I don't see any of the tuning books mentioning that as a relevant performance characteristic. However, I've never been able to move the discussion beyond this point. Can anyone straighten me out on this point or point me to a resource that might be applicable. Our system is Oracle 8.1.6, Compaq Tru64. We use hardware RAID-5 with a battery-backed RAM cache, and have about 3 RAID sets (plus some extra disks for redo logs, etc.), and performance is fine, but I'm always looking as to how we can improve Oracle performance. The application is our corporate ERP system. Dennis Williams DBA Lifetouch, Inc. [EMAIL PROTECTED] -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: Jack C. Applewhite INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: DENNIS WILLIAMS INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: Don Granaman INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list,
Re: RAID system max throughput
This reminds me of my experiences with IBM a couple of years ago. The IBM Shark is a decent piece of storage HW, but does have some funky aspects. Regarding the vendor claims, when I took a close look at their claimed throughput numbers, it became clear that the only way to achieve their claims was to get a 100% hit rate on the Disk Cache. That didn't seem too likely to me. Here's a tidbit: unless they've changed their architecture, IBM Shark only does Raid 5. And you can't get the same number of physical disks in each physical volume. And you must split a PV into at least 2 LV, as the volume manager couldn't make a logical volume as large as the physical volume. Jared Don Granaman granaman@home To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] .comcc: Sent by: Subject: Re: RAID system max throughput [EMAIL PROTECTED] om 12/07/01 03:15 PM Please respond to ORACLE-L Whenever someone (especially a vendor) says something like Don't worry about it..., I worry about it. Who told you that this was simplistic thinking? I've been told similar things a number of times - and proved them wrong in every single case. With hardware RAID, RAID-5 is just as fast as RAID 1+0. With EMC Symmetrix you don't want to stripe. SAME - Just splatter all your files randomly across a monster stripe set using every possible disk. And the ever popular one you are encountering now. A lot of those things are at true - to a point. Beyond that point it matters. Hardware RAID, cache, and such can buy you performance, but there is still some threshold beyond which the old-timey DBA intelligent file placement, striping and such will be necessary. There is a difference between good enough for now and optimal. I would rather build it better from the start, even if I don't need the performance immediately, than wait until its a crisis and only then frantically rebuild everything. See Gaja's paper on RAID at http://www.quest.com/whitepapers/Raid1.pdf . -Don Granaman [OraSaurus] - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Friday, December 07, 2001 4:31 PM Jack - Well, that's what I thought. I could see where the disk would be a lot better about streaming data off the disk if the data was arranged in a favorable manner rather than randomly located. However, I was told that was simplistic thinking and that modern RAID systems are much more sophisticated than that. And I'm willing to concede that a RAID system is more complex than simple drives. I'm just hoping that someone on this list has more experience on the database/hardware interface. Thanks. -Original Message- Sent: Friday, December 07, 2001 3:25 PM To: Multiple recipients of list ORACLE-L Dennis, I'm no RAID guru, but I can sure imagine disk heads thrashing around, trying to satisfy a mix of sequential and random reads and writes, causing the DB to wait, but not getting anywhere near the rated throughput for the RAID controller or channel. Could that possibly be the case? Jack Jack C. Applewhite Database Administrator/Developer OCP Oracle8 DBA iNetProfit, Inc. Austin, Texas www.iNetProfit.com [EMAIL PROTECTED] (512)327-9068 -Original Message- WILLIAMS Sent: Friday, December 07, 2001 10:40 AM To: Multiple recipients of list ORACLE-L Whenever I discuss disk waits with my system
Re: RAID system max throughput
-- DENNIS WILLIAMS [EMAIL PROTECTED] Whenever I discuss disk waits with my system administrator, I always get the reply that the RAID system isn't anywhere near its rated throughput. Maybe I'm wrong, but I don't see any of the tuning books mentioning that as a relevant performance characteristic. However, I've never been able to move the discussion beyond this point. Can anyone straighten me out on this point or point me to a resource that might be applicable. Our system is Oracle 8.1.6, Compaq Tru64. We use hardware RAID-5 with a battery-backed RAM cache, and have about 3 RAID sets (plus some extra disks for redo logs, etc.), and performance is fine, but I'm always looking as to how we can improve Oracle performance. The application is our corporate ERP system. Two things will zap you on device I/O: bandwidth or latency. Most people look at bandwidth -- same basic numbers for both networking and disks. Latency is basically the turnaround time. If you screw up the setup of any I/O system then latency can reduce performance to the point where bandwidth is irrelevant. By analogy, you can put concrete tires on a Porsche and go nowhere also. The real measure of what's going on starts at the O/S level looking at the frequency and duration of proc's in a device wait state (a.k.a. blocked for I/O) on the disks. If this is minimal then forget it. You can also end up with screwy results on large shared disk systems due to competition. SAN's can get placed on overloaded network segments; ERP's can easily get hot-spots from various users colliding. In general RAID5 with a stripe size == system I/O page will perform rather nicely. If your system page sizes vary or the raidset has an offball number of disks (e.g., 6 drives for an 8K page) then you'll take a hit writing extra data to maintain the RAID5 parity. -- Steven Lembark 2930 W. Palmer Workhorse Computing Chicago, IL 60647 +1 800 762 1582 -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: Steven Lembark INET: [EMAIL PROTECTED] Fat City Network Services-- (858) 538-5051 FAX: (858) 538-5051 San Diego, California-- Public Internet access / Mailing Lists To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
